The long tenn goals of this research are (1) to understand how cortical networks represent and :ransform plans into voluntary movement and (2) to use knowledge of cortical representation and;>rocessing for movement to develop neural prosthesis systems that restore function and .ndependence for persons with paralysis. The specific goal of this project is to determine how neural ~ctivity representing reach and grasp is assembled in frontal and parietal cortical networks that :onverge on primary motor cortex (MI). The proposed work combines new dual 96 multielectrode ~ecording and full arm and hand motion capture techniques, as well as novel statistical methods. fhese new tools, which overcome prior limitations in relating cortical ensemble activity to highiimensional voluntary limb actions, will be used to understand how cortical networks transform ntentions into specific arm and hand actions. Aim 1 will extend preliminary findings indicating that vII unifies reach and grasp information locally, rather than segregating limb control into separate arm md hand subdivisions. Aim 2 investigates the way ventral premotor (PMv)-MI networks transform )bject representations into grasp-selective activity patterns within MI. Using free reaching to moving )bjects and grasping of different types, experiments will determine how a precision or power grip is :onstructed from activity related to a viewed object of interest. Aim 3 directly tests the extent to ""hich reaching and grasping functions are segregated into separate or overlapping cortical processing 1etworks and how information is transformed during the dynamic netv.:ork interaction of premotor lOd motor cortex neurons. These shtdies will reveal fundamental principles of cortical information )rocessing that Wlderlie the generation of flexible arm, wrist, and hand actions. In addition, mow ledge about the fonn, location, and generation of cortical movement representations has direct :linical translational relevance to our ongOing human clinical pilot shtdies to develop neural )rosthesis systems. These results will determine the potential to generate a multidimensional control;ignal from one or a combination of cortical areas that could potentially be used to recreate flexible [unreadable]each and grasp actions of robotic anns or paralyzed muscles for persons with severe paralysis.